Prognostic factors after resection of locally advanced non-functional pancreatic neuroendocrine neoplasm: an analysis from the German Cancer Registry Group of the Society of German Tumor Centers

Objective The available literature regarding outcome after pancreatic resection in locally advanced non-functional pNEN (LA-pNEN) is sparse. Therefore, this study evaluates the current survival outcomes and prognostic factors in after resection of LA-pNEN. Materials and methods This population-based analysis was derived from 17 German cancer registries from 2000 to 2019. Patients with upfront resected non-functional non-metastatic LA-pNEN were included. Results Out of 2776 patients with pNEN, 277 met the inclusion criteria. 137 (45%) of the patients were female. The median age was 63 ± 18 years. Lymph node metastasis was present in 45%. G1, G2 and G3 pNEN were found in 39%, 47% and 14% of the patients, respectively. Resection of LA-pNEN resulted in favorable 3-, 5- and 10-year overall survival of 79%, 74%, and 47%. Positive resection margin was the only potentially modifiable independent prognostic factor for overall survival (HR 1.93, 95% CI 1.71–3.69, p value = 0.046), whereas tumor grade G3 (HR 5.26, 95% CI 2.09–13.25, p value < 0.001) and lymphangiosis (HR 2.35, 95% CI 1.20–4.59, p value = 0.012) were the only independent prognostic factors for disease-free survival. Conclusion Resection of LA-pNEN is feasible and associated with favorable overall survival. G1 LA-pNEN with negative resection margins and absence of lymph node metastasis and lymphangiosis might be considered as cured, while those not fulfilling these criteria might be considered as a high-risk group for disease progression. Herein, negative resection margins represent the only potentially modifiable prognostic factor in LA-pNEN but seem to be influenced by tumor grade. Supplementary Information The online version contains supplementary material available at 10.1007/s00432-023-04785-0.


Introduction
Pancreatic neuroendocrine neoplasms (pNEN) are rare pancreatic tumors showing increased incidence over the past few decades. pNEN represent about 7% of all pancreatic tumors (Yadav et al. 2018). According to the European Neuroendocrine Tumor Society (ENETS) guidelines, there are two main forms of pNENs: functional pNEN and nonfunctional (NF) pNEN (Partelli et al. 2017). Unlike functional pNEN, which are hormonally active and are, therefore, often detected early in the course of the disease, NF-pNENs often remain asymptomatic if not incidentally discovered, reaching significant tumor burden until causing symptoms related to the mass effect of the tumor. Therefore, 65% of NF-pNENs are locally advanced or metastatic at presentation (Cloyd and Poultsides 2015).
Despite remarkable advances in the field of oncology and the introduction of novel multimodal treatment protocols, surgical resection remains the only curative treatment of pNEN (Howe et al. 2020;Pavel et al. 2020). The low incidence of pNEN, the wide heterogeneity regarding tumor grade and stage as well as the variety of treatments available today are the main reasons behind the lack of evidence on various aspects of surgical treatment (Jensen et al. 2019).
Until now, the definition of locally advanced pNEN is still controversial and not unified. Most studies included large pNEN with locoregional involvement or oligometastatic disease (Norton et al. 2003;Squires et al. 2020;Fusai et al. 2021). It was first Titan et al. (2020) who defined LA-pNEN as tumors larger than 4 cm (T3/T4) without the presence of distant metastasis. In 2018, the ENETS reported the main areas of unmet needs in the management of patients with functional and non-functional pNEN. Assessment of surgery and identification of prognostic factors in resectable non-metastatic locally advanced pNEN was highlighted as an important aspect requiring further clarification (Jensen et al. 2019).
Due to the low incidence of pNEN in general and this clinical category in specific, multicentric and populationbased registries represent an invaluable source of information. This study aimed to evaluate the clinical outcome and prognostic factors in patients after resection of locally advanced non-metastatic non-functional pancreatic neoplasms (LA-pNEN) using data from population-based cancer registries in Germany.

Patient cohort and study design
This is a retrospective analysis of pooled anonymized cancer registry data. The data were derived from 17 populationbased clinical cancer registries participating in the German Cancer Registry Group of the Society of German Tumor Centers (ADT), covering the time period from 2000 to 2019. The registry data were used according to the regulations of the ADT. The study was approved by the ethics committee of the University of Lübeck, Germany #20-319.

Statistical methods
For statistical analysis, SPSS 26 for Windows (Armonk, NY, USA) was used. For descriptive statistics, absolute and relative frequencies, median and interquartile range as well as Kaplan-Meier estimates and plots were used. Overall survival was calculated as the time period from the date of diagnosis to the date of death. Disease-free survival was defined as the period from the date of tumor resection to the date of local or metastatic recurrence. Statistical testing was performed by Chi squared test, Logrank test and Cox regression using a two-sided significance level of p = 0.05.

Patient cohort
From 2000 to 2019, N = 2446 patients with diagnosis of pNEN were identified. Of those, 280 patients met the inclusion criteria for LA-pNEN. In this cohort, the median age was 63 ± 18 (median ± IQR). There were 155 (56%) patients under 65 years and 122 (44%) females. Data regarding tumor location were available in 207 patients. The tumor was located in the pancreatic head in 106 (51%) patients, in the pancreatic body in 33 (16%) patients, and in the pancreatic tail in 70 (34%) patients. The mean follow-up time was 43 months (95% CI 37-48 months).

Treatment and histopathology
N = 277 patients underwent upfront surgery. Three patients underwent neoadjuvant therapy followed by surgical resection. The type of resection was reported in 183 patients: 91 patients underwent pancreatic head resection, 82 patients underwent distal pancreatectomy, and 10 had a total pancreatectomy.

Recurrence in patients with positive prognostic factors
In the subgroup of n = 71 patients with negative resection margins (R0) and absence of lymph node metastasis and lymphangiosis (N0, L0), only one patient developed local recurrence and 10 patients developed distant metastasis during follow-up (Table S2). The incidence of distant metastasis increased with advanced tumor grade (3% of G1, 25% of G2 and 60% of G3 tumors).

Discussion
The results of this study identified major aspects regarding survival in patients after resection of non-metastatic locally advanced pNEN. First, this study demonstrates that patients with LA-pNEN have favorable long-term 3-, 5-, and 10-year overall survival of 79%, 74%, and 47% respectively, which highlights the importance of offering curative intent surgical therapy at the time of diagnosis. Furthermore, our study demonstrated that positive resection margins (R1/2), tumor location in the pancreatic head or body, and advanced age (> 65 years) were the only independent negative prognostic factors for overall survival in this cohort. Herein, resection margin status represents the only potentially modifiable factor for overall survival in LA-pNEN (Tables 3 and S4). Negative margin status was associated with 5-and 10-year overall survival of 77% and 56%, compared to 61% and 23% in patients with R1/R2 status (Table 2). For this reason, patients with LA-pNEN should be treated in high-volume centers with experience in multivisceral resections and vascular reconstruction to maximize the possibility of R0 resection.
In the current study, 45% of the patients had positive lymph node metastasis. Unlike small pNEN (< 2 cm) were the incidence of LNM in G1 pNEN is negligible compared to G2-G3 tumors (3% vs 16-100%) (Sallinen et al. 2018). In our study, lymph node metastasis was present in 37% of G1, 46% of G2 and 57% of G3 of LA-PNEN (Table S1). Interestingly, lymph node metastasis was not associated with overall survival. However, lymph node metastasis and lymphangiosis, as well as advanced tumor grade, were associated with shorter disease-free survival. Furthermore, we identified G3 pNEN and lymphangiosis as the only independent negative prognostic factors of disease-free survival in LA-pNEN. Since higher tumor grade was also associated with positive resection margin, we can suggest that the grade of differentiation relates to lymphatic tumor spread and locoregional tumor dissemination, therefore, leads to positive resection margin and recurrence. This implies also that lymphadenectomy should be carried out in LA-pNEN to achieve local control and avoid tumor recurrence after curative resection. Another observation is that, after lymphadenectomy and resection of positive lymph nodes, recurrence and eventually death occurs relatively late due to the slowgrowing nature of pNEN. Therefore, long follow-up times are needed to analyze overall survival, but tumor biology is probably better reflected in disease-free survival at earlier time points. This might explain the observed differences in prognostic factors of OS and DFS. In the subgroup analysis of patients with favorable prognostic factors (R0, N0, L0, n = 71), only one patient developed local recurrence and 10 patients developed distant metastasis during follow-up (Table S2). Almost all recurrences occurred in G2 and G3 tumors. This means, cure can be achieved in G1 LA-pNEN in the presence of favorable histopathological factors such as L0, N0, and margin negative resection (R0). As such, resection margin, grade of differentiation, lymph node metastasis and lymphangiosis may also serve as prognostic factors to identify patients with a high-risk for metastasis and local recurrence.
LA-pNENs represent a subgroup of patients with a high risk for vascular involvement and multivisceral resection during exploration. Fusai et al. reported that 25% of T3-T4 pNEN required portal vein reconstruction, compared to only 3% in patients with T1-T2 pNEN in the same cohort   (Fusai et al. 2021). Similarly, Titan et al. reported that 25% of LA-pNENs were associated with vascular invasion in the radiological studies, and 17% required eventually vascular reconstruction (Titan et al. 2020). They also reported a very high 10-year OS of 91% compared to our results of 47%. This could be due to the differences in multimodal therapy regimes, differences in the age groups (57 vs 63 years), and tumor grading (G1 54% vs 39%, G2 30% vs 47%, and G3 1% vs 14%), and more importantly, because they included functional pNEN such as insulinomas (n = 9) in their cohort, which are known to have excellent prognosis after curative resection (Roland et al. 2012;Kasumova et al. 2017). Unfortunately, the details of surgical resection (i.e., extent of multivisceral resection, extent of lymphadenectomy) and reconstruction are not available in the cancer registry data used in this study. However, due to comparable R0 resections, we can assume that similar surgical approaches (89% vs 84%) have been adopted in the treatment of LA-pNEN. Another possibility is to offer patients with LA-pNEN neoadjuvant treatment, analogous to patients with borderline resectable pancreatic ductal adenocarcinoma (BR-PDAC). Although this treatment strategy is not well established in pNEN, several agents have demonstrated promising results, including systemic chemotherapies (Prakash et al. 2017;Squires et al. 2020), PRRT (Parghane et al. 2021;Zanata et al. 2021), and targeted therapy such as everolimus (Sato et al. 2017). These protocols could offer a solution to increase resectability even in initially unresectable pNENs.
Several limitations should be considered when interpreting our results. Although the registry data are the result of multi-institutional collaboration, possible inconsistency in patient selection, surgical experience, pathologic assessment, and reporting likely existed. Moreover, the German Clinical Cancer Registry Group collects data from several regional clinical cancer registries implicating data entry by different people in variable quality. Furthermore, the registry data did not include detailed information regarding treatment schemes, timing, and completion of these protocols in pNENs or surgery-related complications, such as pancreatic fistula or postoperative morbidity. In addition, information regarding Ki67% or the differentiation between pNEN G3 or NEC was not available in queried dataset. Nevertheless, this is the largest and the first registry-based analysis of treatment and survival outcomes of locally advanced non-functional pNEN without distant metastasis. Moreover, we demonstrated the favorable prognosis of LA-pNEN and identified the involvement of resection margins as the only modifiable prognosticator in this subgroup of patients. Future studies might address the impact of neoadjuvant therapy on resection status and the relevance of circumferential resection margin on prognosis in LA-pNEN.

Conclusion
Resection of locally advanced pNEN with negative resection margins is associated with favorable 5-and 10-year overall survival (79% and 47%). The resection margin was the only potentially modifiable prognostic factor for LA-pNEN. Positive lymph node metastasis was present in almost half of the patients; therefore, resection with curative intent and radical lymphadenectomy should be routinely performed in LA-pNEN. Patients after resection of LA-pNEN with R0 N0 L0 G1 histopathology might be considered cured, while patients not meeting these criteria to be categorized as highrisk patients for local recurrence and distant metastasis.

Conflict of interest
The authors declare no conflict of interest.
Institutional review board statement This study received approval from the Ethics Committee of the University of Lübeck (ethics committee protocol #20-319).
Informed consent statement Patient consent was waived due to the analysis of anonymized data.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.